‘Economic Policy in a European Monetary Union’ was the title of a joint conference organized by CEPR and the European Forum of the European University Institute, and held in Florence on 5/6 December 1996. The conference, which was organized by Michael Artis, Ramon Marimon (both European University Institute, Florence, and CEPR) and Francesco Giavazzi (IGIER, Università Bocconi and CEPR), formed part of the 1996/97 European Forum on ‘The Political Economy of an Integrated Europe’. It was supported by the Margaret and Richard Merrell Foundation.

The main theme of the seven papers presented at the conference was the design of fiscal and monetary policies for EMU. In addition, a panel session was dedicated to current policy issues, especially the controversial proposals for a ‘stability pact’ to constrain fiscal policy in EMU and the procedure for selecting countries for EMU membership.

On both fiscal and monetary policies, much of the debate concentrated on the relative merits of policy rules and constraints versus the need for macro- economic stabilization, i.e. the apparent trade-off between credibility and flexibility. This was most evident in the panel discussions on the proposed stability pact. For some, the pact was a useful device to secure fiscal consolidation, which was required independently of EMU; for others, it posed dangers of adverse effects on output, suppression of automatic stabilizers and imposition of excessive adjustment beyond what was economically desirable and politically acceptable. Proponents rejected the notion of a trade-off between fiscal discipline and flexibility. On the contrary, regaining sound and sustainable budget positions was a precondition for automatic stabilizers to function effectively.

The stability pact and the Maastricht convergence criteria were seen mainly as disciplinary devices for fiscal policy, to the neglect of other issues. There was some recognition that budget rules were inferior to fully fledged coordination of fiscal policies, but no argument as to to how the latter might be achieved under EMU. The important problem of fiscal and monetary policy coordination, including the potential roles for the criteria and the stability pact, also was left largely unaddressed. Thus, the conference papers dealt with either fiscal or monetary policy issues, but did not seek to investigate links between the two. This would seem an important area for further research.

The opening paper by Paul De Grauwe (Katholieke Universiteit Leuven and CEPR) addressed the important practical issue of ‘How to Fix Conversion Rates at the Start of EMU’ by comparing two basic possibilities. Either (1) the initial currency conversion rates already will have been announced by policy-makers when they decide in early 1998 which countries will participate, or (2) the markets will decide. A version of the latter option would be to fix conversion rates as the average of previous market rates, dubbed the ‘Lamfalussy rule’, after the president of the European Monetary Institute.

De Grauwe used a four-period model for illustration, where the current exchange rate is a function of fundamentals and of expectations of future exchange rates. The model suggested a disturbing result. Unlike the case of exchange rates (credibly) fixed ex ante, and unlike pure floating, the Lamfalussy rule produced an amplification effect of shocks in fundamentals, leading to greater exchange-rate instability. Therefore, a prior announcement of conversion rates was to be preferred. To enhance the credibility of these conversion rates, De Grauwe proposed the prohibition of unilateral parity adjustments, which was already in the Maastricht Treaty. Furthermore, ‘institutional front-loading’ should require monetary policy cooperation by 1998, for example by national central banks targeting European money supply.

Luigi Spaventa (Università degli Studi di Roma ‘La Sapienza’ and CEPR) commented that the ‘market view’ was that the Lamfalussy rule was an attractive compromise between the extremes of pure floating and early fixing of rates, and that the relevant trade-off was between flexibility and credibility. He warned that policy-makers should decide only the central bilateral national exchange rates and not the rates with the Euro, but the ‘ERM-II’ rates of EMU outsiders also should be fixed at the same time. De Grauwe responded that fixing Euro rates instead of bilateral rates was technically possible, as long as one country was willing to have its currency act as the numeraire, but this would be difficult politically.

Peter McAdam (University of Strathclyde), in a joint empirical study with Andrew Hughes Hallett (University of Strathclyde and CEPR), asked ‘Is a Stability Pact the Answer to Europe’s Coordination and Fiscal Discipline Problems?’ The authors used the IMF’s Multimod model to simulate the impact of the proposed stability pact from 1999 onwards. This implied budget deficits no greater than 3% of GDP if countries wished to avoid paying a fine of 0.25% of GDP for each percentage point in excess of the target. While there were some positive effects from fiscal contraction on output and employment, the net effect was clearly negative, most dramatically so for Italy, which would be condemned to permanent recession. The authors also argued that any attempt to seek deficit reduction by cutting expenditure – rather than via adjustment of taxes – would fail, because the expenditure multiplier was close to unity. Thus, both the deficit and output would be affected in the same proportion, leaving their ratio unaltered. The authors also warned against an ‘instability’ pact that could lead to an unbalanced policy mix, and to perverse effects on fiscal consolidation, as recession would increase the deficit ratios, and low growth at high real interest rates could cause debt ratios to explode.

Mark Salmon (European University Institute and CEPR) emphasized the importance of seeking quantitative answers to policy questions even if simulation studies could be easily criticized. Even so, he would have preferred an explicit analysis of intertemporal optimization and policy cooperation. That would have allowed clearer evaluation of the losses arising from implemention of ad hoc intermediate targets, such as the Maastricht criteria and the ‘instability’ pact. A particular question was why all countries should be subjected to the same rules, which would have different effects across countries. Flemming Larsen (IMF) noted that other simulations at the IMF with the same macromodel had produced results somewhat at odds with those of the paper’s authors. In particular, results depended critically on the benchmark adopted and, in the absence of any fiscal rules, a continuation of past and present fiscal policies would be neither optimal, nor perhaps sustainable. Paul De Grauwe noted that the large and sustained inflation differentials (and therefore price level divergences) predicted in the simulations were not consistent with the inflation process in a monetary union.

The Round Table on ‘Convergence Towards EMU and the Stability Pact’ was opened by Ad van Riet (European Monetary Institute), who summarized the essential elements of the EMI’s convergence report. In evaluating whether countries had achieved a sufficient degree of sustainable convergence, the EMI would follow four guiding principles with respect to the Maastricht criteria: interpretation for the 1998 report would be strict, part of a coherent and integrated package, based on realized 1997 data, and transparent. Progress on the inflation, interest-rate and exchange-rate criteria had been noteworthy in 1996, but on the fiscal side, most countries had not yet achieved sustainable medium-term positions. A ‘sixth convergence criterion’ in the treaty also called for the achievement of central bank independence by 1999 at the latest for all countries without an opt-out.

For Antonio José Cabral (European Commission, DG II) the proposal for a stability pact rested first, on the general case for fiscal consolidation and, second, on the pragmatic need to help countries fulfil their Maastricht Treaty obligation to avoid excessive deficits. On the first count, reduced deficits were needed in any case to preserve the viability of social security and other vital government spending, without increasing the tax burden. In this context, deficit rules, even if they appeared ‘arbitrary’, would lead to better results than fiscal discretion. On the second count, anything that – like the stability pact – strengthened the credibility of existing treaty provisions, particularly the excessive-deficit and no-bailout clauses in Article 104c, should be welcomed. The aim of the pact was to speed up and operationalize implementation of deficit reductions, via prevention, surveillance and deterrence.

Peter Kenen (Princeton University) supported the view that the proposed stability pact was dangerous. A careful reading of Article 109j revealed that fulfilment of the convergence criteria was neither a necessary nor a sufficient condition for EMU membership. Another ambiguity in the treaty concerned the possibility of a delay of the starting date. EMU would start on 1 January 1999 only if no other date had been set by the end of 1997. It was therefore conceivable that the Council could decide in 1997 to delay to a ‘certain’ later date. The stability pact would accelerate imposition of penalties for excessive deficits, and therefore ‘short-circuit’ the operation of automatic stabilizers during a recession. Kenen also advised against accepting a pact as a trade-off in return for relaxed entry conditions because of the need to preserve fiscal flexibility in EMU.

Flemming Larsen was more sanguine about the stability pact. He estimated that ten countries could pass the crucial deficit criterion. For five of these, however, there was still considerable doubt, owing to the great uncertainties about next year’s (1997) cyclical position and the degree to which ‘creative accounting’ would be acceptable. The fixation on the 3% deficit criterion was misguided in these circumstances. Rather than postponing EMU, a flexible interpretation of the criterion should focus on the structural deficit, and a stability pact agreed to regain flexibility for the future. Given that fiscal adjustment was required in any case, its short-run costs would be reduced the sooner EMU happened and a pact was agreed, since risk premia and interest rates should fall in response. While the pact could provide a solution to the dilemma in 1999, in the longer run the true microeconomic aspects of fiscal problems would have to be addressed. These included inflexible labour markets, excessive taxation and unfunded pension liabilities.

Michael Artis also saw merit in the pact. Discussion of the fiscal conditions had centred on the deficit, rather than the debt, criterion. The ability to issue and service debt, however, depended on the power to raise taxes, and on the power to print money. The former diminishes with increasing factor mobility in Europe, leading to inter-state tax competition; the latter is taken away completely from national governments. This meant that debt runs could not be ruled out, and they would be hard to control in EMU. It followed that the Treaty’s no-bail-out clause may be neither very smart nor very credible. The stability pact was seen as the key issue for the fate of EMU and a necessary condition to win over a sceptical German public. The main fault with the convergence criteria was that they were not state-contingent and, therefore, they failed in their purpose of testing for ‘sustained dedication to stability culture’. Because of creative accounting, and with even Germany experiencing great difficulties, the criteria had become ‘non-revealing’ as entry conditions. Here, the stability pact could perhaps compensate for the failure of the criteria, and provide conditions for the long-run stability of EMU.

In the open floor discussion, Paul De Grauwe agreed with Peter Kenen’s criticism of the stability pact. The danger was mainly political, in that ‘Europe’, with its debatable democratic credentials, would be perceived as imposing unpopular policies – especially in recessions – on member states. Luigi Spaventa thought that enforcement of the pact would be weak given the time lags involved and especially the availability of annual data. Moreover, since different levels of debt – whether short-term or variable-rate – led ‘symmetric shocks’ and monetary policy to have an asymmetric impact, Spaventa proposed an additional debt-composition entry requirement for EMU. Alasdair Smith (University of Sussex and CEPR) asked what would happen if Italy adopted the Deutschmark unilaterally, rather than merging currencies multilaterally in EMU. De Grauwe argued that Italy would impose externalities on its partners, whether or not it took part in EMU. If Italy stayed out, the externalities from a weak lira exchange rate would hurt EMU-insiders. Externalities did not, therefore, afford a good argument for excluding Italy.

Harald Uhlig (CentER, Tilburg University, and CEPR) examined ‘Long-Term Debt and the Political Support for a Monetary Union’. The central idea was that a decision to join (a low-inflation) monetary union would lead to a redistribution of income from debtors to creditors of long-term nominal fixed-rate debt of (previously) high-inflation countries. Creditors should therefore support entry into monetary union and the size of their windfall gain could be very significant. But there was an offsetting ‘tax effect’, since their citizens also would have to pay higher taxes to foot increased real debt payments on government bonds. Different options for strategic debt management to influence political support were explored in the paper, including issuing debt in foreign currencies. The main policy conclusion suggested that governments wanting political support for EMU should not issue long-term bonds in the home currency but in a foreign currency.

Robert Waldman (European University Institute) wondered why citizens might vote against EMU while their government was in favour. One explanation could be ‘dynamic inconsistency’ of voter preferences occasioned by unhappiness with EMU’s short-term costs as against its long-term benefits. He identified three possible sources of short-run costs. First, surprise deflation could lead to recession in the presence of nominal contracts; second, high marginal tax rates to pay for EMU entry were suboptimal from a tax-smoothing perspective; third, and perhaps most important, people may be irrational, i.e. they may not understand that bonds are not net wealth if bondholders are also taxpayers. Paul De Grauwe wondered how to square the paper with the widely held view that high-inflation countries would obtain lower real interest rates from EMU, rather than surprise deflation leading to higher real interest rates as in the model. Luigi Spaventa responded that the decline of real interest rates from EMU applied to newly issued debt, as opposed to the stock of existing debt. The net effect would depend on the precise maturity structure of existing debt.

Typhon Kollintzas (Athens University of Economics and Business and CEPR) looked at ‘Fiscal Policy in an Economic Union’ – a joint work with Apostolis Philippopoulos (Athens University of Economics and Business and University of Essex) and Vanghelis Vassilatos (IMO, Athens). The central idea of their paper was that economic union would increase spillovers from tax policies, and intensify tax competition among member states. In a dynamic game, national governments would act as Stackelberg leaders vis-a-vis the private sector, but play Nash strategies against the other governments. A supranational government was also introduced, and both normative and positive issues of strategic interaction of fiscal policies were highlighted.

Robert Kollmann (Université de Grenoble and Université de Montreal) thought the paper represented an ambitious attempt to extend the existing two-country real business cycle literature, which had treated government behaviour as exogenous. For a meaningful analysis of fiscal policy, however, it would be necessary to move beyond the representative-agent framework, to allow governments to issue debt and introduce different taxes on labour and capital. To increase the empirical relevance of this type of model, capital adjustment costs also would be important. Alasdair Smith asked whether EMU would have an effect on the mobility of factors and increase tax competition over and above the impact of the single market.

The paper by Pedro Teles (Banco de Portugal and Universidad Catolica Portugesa) explored ‘The Virtues of a Parallel Currency System’, using a neoclassical two-country public-finance model. The countries differed in their (implicit) preferences for the inflation tax, and there were externalities in seigniorage collection, leading to an inflationary bias. National, single and parallel currency regimes were confronted in terms of their coordination properties and, hence, their welfare implications. A parallel currency, issued by a common monetary authority interacting strategically with national authorities, is shown to dominate a single currency. The parallel currency puts a ceiling on inflation in both countries, while maintaining a degree of freedom to account for different preferences. A single currency cannot do this and so could make countries worse off than when separate national currencies are retained.

Ramon Marimon suggested that a more complete picture also would have to address the dangers of parallel currencies. The idea of having different currencies compete had a long tradition, but could induce exchange-rate instability – as the experience of dollarization in Latin America had shown – or even lead to a complete currency collapse. The latter might not necessarily be welfare-decreasing, but would raise distributional issues as nominal assets held in the currency became worthless. Analysis of parallel currency regimes remained relevant for EMU for the transition period 1999–2002, as long as national currencies circulated alongside the Euro and monetary union remained reversible. Moreover, countries left outside EMU – in particular, the Eastern European countries – might choose to adopt the Euro as a parallel currency. Paul De Grauwe thought that since currency competition implied floating exchange rates, it could not be reconciled with monetary unification, but would represent a step backwards from the exchange-rate stability achieved in Europe. Peter Kenen objected that the public-finance approach to inflation as an optimal tax was not a good model of the money supply process. Furthermore, the costs of a currency collapse were enormous, because it usually involved a banking collapse and subsequent government bailout. Hubert Kempf (Université de Paris I) suggested that a comparison of different currency regimes should have a role for transaction costs.

The paper by Sylvester Eijffinger (Humboldt Universität Berlin and CentER, Tilburg University) investigated ‘The Optimal Conservativeness of the European Central Bank’. He stressed the importance of distinguishing between conservatism and independence, which are lumped together in many indices of central bank independence. While conservatism concerned central bank preferences, independence captured the balance of power between the government and the central bank, i.e. the degree to which the bank can enforce its objectives on monetary policy choices. Both parameters were explained theoretically and empirically as a function of four economic and political determinants, namely the natural rate of unemployment, society’s output-inflation preferences, the variance of productivity shocks, and the benefits of unanticipated inflation. Empirical evidence from 12 EU countries revealed a positive relationship between the estimated degree of optimal conservatism (before EMU) and long-term interest rates. Countries like Belgium, Finland, Italy and Spain would gain a lot from EMU in terms of independence, and therefore should be included from the start. By contrast, Denmark, Ireland, Sweden and the UK should not join the core group of EMU (Austria, France, Germany and the Netherlands).

Berthold Herrendorf (University of Warwick) raised several objections to the model and the general approach in the literature on central bank independence. In particular, it seemed odd that independent central bankers should somehow be able to choose their degree of conservatism (i.e. their own preferences) optimally. For the empirical estimates, both the exchange-rate regime and the presence of output persistence and policy lags should affect the results. More generally, it was not at all clear why an independent central bank should target employment above the natural rate and therefore suffer from a credibility problem.

Ramon Marimon wondered whether EMU member countries would not appoint ECB representatives to defend national interests rather than optimally conservative ones. Luigi Spaventa regretted that the model was missing a transmission mechanism for monetary policy, which would vary across countries because of institutional differences.

In the final paper on ‘Market Discipline and Monetary Policy’, Carl Walsh (University of California at Santa Cruz) advanced the possibility that an increased role for market expectations could lead to a reduction in the inflation bias of monetary policy. Inflation was modelled as a function of the output gap and expected future inflation. Whereas a ‘strong’ policy-maker could commit to the path of future inflation, a ‘weak’ one could not, and the public (market) was uncertain about which type was in office. In a two-period model an interesting result was the possibility of a first-period expansion when a strong type was in office. The expectation of lower future inflation led to an expansion in the current period, potentially offsetting the negative effect from surprise deflation. The effect from expectations on future inflation increased the discipline of weak policy-makers, i.e. added to their incentive to behave as if they were strong. Adding market expectation effects therefore was similar to increasing the discount factor – i.e. the weight placed on the future – in the policy-maker’s objective function.

Giorgia Giovannetti (Università di Firenze) welcomed the idea of modelling forward-looking expectations. Empirically, the latter could be seen in the swift reaction of the interest-rate term structure to news. She suggested exploration of the implications of extending the models beyond two periods, and consideration of policy-maker types differing in their preferences rather than their commitment ability. Luigi Spaventa asked what institutional factors would determine the degree to which current inflation reacted to market expectations, and he suggested three possibilities: the degree of indexation, the development of financial markets and the degree of openness of the economy. Ramon Marimon proposed an analogy with learning models which suggested that, if great weight was placed on new information, instability would increase.